Contemporary Clinical Trials 41 (2015) 9–16

Contents lists available at ScienceDirect

Contemporary Clinical Trials journal homepage: www.elsevier.com/locate/conclintrial

Hepatitis C vaccine clinical trials among people who use drugs: potential for participation and involvement in recruitment April M. Young a,b,⁎, Dustin B. Stephens c, Hanan A. Khaleel a, Jennifer R. Havens b,c a b c

Department of Epidemiology, University of Kentucky, Lexington, KY, United States Center on Drug and Alcohol Research, University of Kentucky, Lexington, KY, United States Department of Behavioral Science, University of Kentucky, Lexington, KY, United States

a r t i c l e

i n f o

Article history: Received 10 October 2014 Received in revised form 20 December 2014 Accepted 22 December 2014 Available online 30 December 2014 Keywords: Clinical trial Drug users Hepatitis C Vaccine Injection drug use Social network

a b s t r a c t Candidate prophylactic HCV vaccines are approaching phase III clinical trial readiness, yet little is known about the potential for participation among target groups or innovative ways to promote enrollment within ‘hard-to-reach’ populations. This study describes HCV vaccine trial participation willingness among a high-risk sample of people who use drugs and their willingness to assist researchers by promoting the trial among peers. Willingness to participate in and encourage peers' participation in an HCV vaccine trial was assessed among injection and non-injection drug users enrolled in a cohort study in Kentucky using interviewer-administered questionnaires (n = 165 and 415, respectively, with willingness to participate assessed among HCV-seronegative participants only). Generalized linear mixed models were used to determine correlates to being “very likely” to participate or encourage participation in a trial. Most reported being likely to participate or encourage participation in a vaccine trial (63% and 87%, respectively). Men were significantly less likely to report willingness to encourage others' participation, while willingness to encourage was higher among lower income, HCV-seropositive, heroin-using, and methamphetamine-using participants. Unemployment, lesser education, receipt of financial support from more peers, and nonmedical prescription drug use were positively associated with willingness to participate. Differential enrollment in HCV vaccine clinical trials by socioeconomic status may occur, underscoring ethical considerations and need for avoiding coercion. Notably, the data suggest that a peer-driven approach to promoting trial participation among people who use drugs could be feasible in this population and that HCV-seropositive individuals and women could be especially instrumental in these efforts. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Over 450,000 people worldwide die annually from liver disease caused by hepatitis C virus (HCV) [1]. Mortality resulting from HCV in the United States (US) now exceeds Abbreviations: HCV, hepatitis C virus; HIV, human immunodeficiency virus; IDU, injection drug use; PWID, people who inject drugs; WEP, willingness to encourage participation in an HCV vaccine clinical trial; WTP, willingness to participate in an HCV vaccine clinical trial. ⁎ Corresponding author at: Department of Epidemiology, University of Kentucky College of Public Health, 111 Washington Avenue, Office 211C, Lexington, KY 40536, United States. Tel.: +1 859 218 2090; fax: +1 859 257 8811. E-mail address: [email protected] (A.M. Young).

http://dx.doi.org/10.1016/j.cct.2014.12.015 1551-7144/© 2014 Elsevier Inc. All rights reserved.

that of HIV [2]. Transmission persists in populations at risk for parenteral exposure to the virus, particularly via injection drug use (IDU) [3]. There is evidence of rising incidence among young people who inject drugs (PWID) in the US [4,5], where 50 to 80% of PWID are infected with HCV [6]. Given the high prevalence and infectiousness of HCV via parenteral transmission [7], behavioral interventions to reduce HCV transmission should be supplemented by biomedical approaches to prevention, potentially including prophylactic vaccination [8,9]. HCV vaccination among PWID could be a cost-effective strategy for decreasing future HCV-related healthcare burden [9–11]. Several candidate prophylactic HCV vaccines are nearing readiness for phase III trials [12–16] and research supports the

10

A.M. Young et al. / Contemporary Clinical Trials 41 (2015) 9–16

feasibility of a large, multi-center HCV vaccine trial among PWID [17]. PWID [3], as well as those sharing straws for snorting [18,19] are at risk for HCV acquisition and could be an appropriate study population for trials; however, recruitment may be challenging [20]. Given the importance of involving atrisk populations in HCV vaccine trials, identification of factors associated with willingness to participate (WTP) is vital, as is research investigating novel approaches to recruitment. Few studies have examined WTP among HCV-seronegative PWID. Extant research indicates that trial-related characteristics (e.g., safety, privacy, efficacy, trial duration, compensation, vaccine administration method), as well as altruism, peer support, mistrust, confidentiality concerns, and comprehension of the concept of clinical trials, vaccines, and/or HCV may affect WTP [21–25]. In addition, two qualitative studies reported financial compensation as a key motivator for trial participation, along with altruistic motives and positive peer communication and support [23,24]. Despite evidence that peer communication and support can play a role in promoting participation in HCV and HIV clinical trial research [20,23,26], no study to date has explored the feasibility of involving peer-promotion in HCV trial recruitment among drug users. The purpose of this study was to examine drug users' willingness to participate in and encourage their drug-using peers to participate in a clinical trial for a prophylactic HCV vaccine in the context of participants' injecting networks. This study was conducted in rural Appalachian Kentucky, a region with high prevalence of HCV among PWID [27,28] in the state reporting the highest number of acute HCV infections in the United States [29]. The incidence rate among PWID in this region (approximately 14 per 100 person–years) [30] is near the primary infection rate estimated in a recent study to be necessary to adequately power a trial evaluating efficacy of a highly efficacious vaccine designed to prevent chronic HCV [31]. In an economically deprived area with limited healthcare access, under-resourced social service structure [32–34], and prohibition of needle/syringe provision (NSP) [35,36], the future HCVrelated healthcare burden is likely staggering. Thus, the region would greatly benefit from an efficacious vaccine and present a prime population for HCV vaccine research.

Following their 24-month interview, participants (n = 433) were invited to complete an additional intervieweradministered questionnaire that examined their attitudes toward HIV and HCV vaccines and clinical trial participation. All invited participants provided informed, written consent to participate and were compensated $35. The University's Institutional Review Board approved the protocol. 2.1.1. HCV vaccine clinical trial participation Preceding the questions about HCV vaccine clinical trial participation, participants were read the following script: “Hepatitis C is a virus that can be transmitted from person to person through sharing drug equipment such as syringes, cookers, cottons, and rinse water. Scientists are working on developing a vaccine that would prevent people from getting hepatitis C. It would not cure hepatitis C, it would only stop people from getting it.” Then, willingness to participate in a clinical trial for a prophylactic HCV vaccine was assessed with the following: “Before the hepatitis C vaccine can be approved for use for everyone, researchers must study the vaccine in clinical trials. In a clinical trial, researchers give volunteers an experimental vaccine to study the effects of the vaccine. If there was a clinical trial on hepatitis C vaccines in this community, how likely would you be to volunteer to be in it?”, followed by a 4-point Likert scale ranging from ‘very unlikely’ to ‘very likely’. Participants were also given the option to indicate that the question did not apply to them because they already were infected with HCV. Participants, including those with HCV, were then asked, “How likely would you be to encourage the people you use drugs with to be in a clinical trial for a vaccine against hepatitis C?”, with responses arranged on the same 4-point Likert scale described above.

2.1. Sample

2.1.2. Demographic and behavioral characteristics Participants also provided demographic data, including age, gender, race, marital status, education, employment, total monthly income, health insurance status, self-reported health, and recent (past 30 day) incarceration. Income was recoded into quartiles and treated as ordinal in analysis due to positive skew of the continuous distribution. Behavioral data collected from participants included recent (past 6-month) alcohol and illicit drug use (substances listed in Tables 1 and 2), injection drug use, use of an unclean needle, having gave/loaned/sold a used needle to someone, and unprotected sex with PWID.

The data used for this analysis were collected during the 24-month assessment of the longitudinal Social Networks among Appalachian People study (described in detail elsewhere [37]). Participants (n = 503) were recruited using respondentdriven sampling and, to be eligible for participation, were required to be 18 or older, reside in Appalachian Kentucky, and to have used prescription opioids, heroin, crack/cocaine, or methamphetamine “to get high” in the prior 30 days. Questionnaires and HCV testing were administered by trained, community-based staff approximately every 6 months and the follow-up rate at the time of data collection for the present analysis exceeded 90%. Participants who tested antibodypositive in a previous follow-up assessment were not re-tested at subsequent visits. Participants were tested using the OraQuick (Bethlehem, PA) [38] Rapid-HCV test and were provided with post-test counseling tailored to their study result.

2.1.3. Network characteristics The questionnaire also elicited network data (described in detail elsewhere [39]). Briefly, each participant, or ‘ego’, gave the first name and last initial and demographic information (e.g., age and gender) of up to eight individuals, or ‘alters’, from/with whom they had received social support, used drugs (excluding alcohol and marijuana), and engaged in sex during the past 6 months. Of note, none of the participants named more than seven drug or sex network members and only one named eight social support network members; therefore, it is unlikely that limiting participants to eight named partners per network restricted data on network size. Respondents were then asked with whom of the individuals they listed had they shared drugs (frequency on 5-point Likert scale), shared injection equipment (sum of two 5-point Likert scales on which participants rated the frequency of needle and cooker

2. Materials and methods

A.M. Young et al. / Contemporary Clinical Trials 41 (2015) 9–16

11

Table 1 Univariate correlates to being ‘very likely’ to encourage drug-using partners to participate in a clinical trial for an HCV vaccine (n = 415).a Characteristicb

Willingness to encourage trial participation — n (%) Not very likelyc (n = 216)

Demographic and serostatus Male — n(%) White — n(%) Age — mean (SD) Total Monthly Income First quartile ($0–$200) Second quartile ($201–$698) Third quartile ($699–$1100) Fourth quartile (≥$1101) Education (years) — mean (SD) Married — n(%) Uninsured — n(%) Unemployed — n(%) Self-report of being in good health — n(%) Incarcerated (past 30 days) — n(%) HCV-seropositive — n(%)

Very likely (n = 199)

OR (95% CI)

p-value b0.001⁎⁎ 0.116 0.088 0.049⁎

149 (69.0) 199 (92.1) 35.8 (8.8)

79 (39.7) 190 (95.5) 34.7 (8.2)

0.16 (0.08–0.34) 2.52 (0.80–8.00) 0.97 (0.95–1.00)

51 (23.7) 46 (21.4) 55 (25.6) 63 (29.3) 11.22 (1.97) 54 (25.0) 135 (62.5) 82 (38.0) 132 (61.1) 56 (25.9) 121 (56.0)

65 (32.7) 51 (25.6) 42 (21.1) 41 (20.6) 11.24 (1.81) 55 (27.6) 140 (70.4) 82 (41.2) 168 (84.4) 59 (29.7) 131 (65.8)

2.82 (1.32–6.02) 2.14 (0.98–4.66) 1.48 (0.69–3.15) Reference 1.03 (0.90–1.19) 1.24 (0.69–2.24) 1.50 (0.86–2.62) 1.00 (0.59–1.70) 5.49 (2.69–11.23) 2.11 (1.00–4.47) 2.21 (1.27–3.83)

0.646 0.470 0.154 0.993 b0.001⁎⁎ 0.050 0.005⁎⁎

Substance use in past 6 months Alcohol — n(%) Alcohol to intoxication — n(%) Nonmedical use of prescription drugs — n(%) Heroin — n(%) Methamphetamine — n(%) Cocaine — n(%) Crack — n(%)

92 (42.6) 68 (31.5) 186 (86.1) 6 (2.8) 13 (6.0) 24 (11.1) 5 (2.3)

82 (41.2) 51 (25.6) 166 (83.4) 16 (8.0) 22 (11.1) 27 (13.6) 9 (4.5)

1.03 (0.61–1.74) 0.58 (0.33–1.01) 0.86 (0.42–1.76) 4.25 (1.13–15.94) 2.89 (1.05–8.00) 1.20 (0.52–2.79) 3.15 (0.66–15.09)

0.921 0.055 0.678 0.032⁎ 0.041⁎ 0.666 0.152

Risk behaviord Injected drugs — n(%) Injected with unclean needle — n(%) Distributed unclean needlee — n(%) Shared injection equipmentf — n(%) Bleached needles — n(%) Unprotected sex with PWID — n(%) Snorted drugs (past 30 days) — n(%)

70 (32.4) 13 (6.0) 7 (3.2) 24 (11.1) 15 (6.9) 36 (16.7) 105 (48.6)

70 (35.2) 20 (10.1) 9 (4.5) 29 (14.6) 18 (9.1) 45 (22.6) 77 (38.7)

1.14 (0.66–1.96) 1.75 (0.67–4.57) 1.01 (0.27–3.77) 1.22 (0.56–2.65) 1.30 (0.52–3.22) 1.80 (0.95–3.42) 0.67 (0.39–1.17)

0.642 0.252 0.993 0.619 0.576 0.073 0.158

0.28 (0.84) 0.58 (1.05)

0.17 (0.65) 0.58 (1.13)

0.84 (0.59–1.21) 1.05 (0.82–1.35)

0.356 0.695

0.26 (0.78)

0.21 (0.69)

0.87 (0.59–1.28)

0.482

0.59 (2.94)

0.38 (1.85)

0.91 (0.78–1.05)

0.179

1.46 (1.68) 3.83 (4.52) 1.74 (1.23) 0.58 (0.69)

1.24 (1.66) 3.16 (4.60) 1.62 (1.22) 1.05 (0.97)

0.94 (0.79–1.12) 0.97 (0.91–1.03) 0.91 (0.72–1.16) 2.56 (1.62–4.04)

0.501 0.354 0.451 b0.001⁎⁎

Network characteristics (n = 356)g Number of injection partners — mean (SD) Number of people in personal network who inject — mean (SD) Number of people to whom participant talks about injection-related risk reduction — mean (SD) Frequency of sharing injection equipment with all network members — mean (SD) Number of people with whom participant uses drugs — mean (SD) Frequency of drug sharing with all network members — mean (SD) Number of network members providing social support — mean (SD) Number of network members providing financial support — mean (SD)

PWID: person who injects drugs; OR: odds ratio; CI: confidence interval; SD: standard deviation. ⁎ p-value b 0.05. ⁎⁎ p-value b 0.01. a Does not include those who reported that the question was not applicable (n = 18). b All categorical variables (indicted with a ‘n(%)’) were dichotomous. c Includes responses “very unlikely”, “unlikely”, and “likely”. d The recall period for risk behaviors listed was 6 months with the exception of snorting, which was assessed based on the past 30 days. e Sold, loaned, or gave needle to someone after using it. f Cookers, cottons, and/or rinse water. g 59 participants reported no network members.

sharing), discussed risk reduction (i.e. bleaching and/or not sharing injection equipment), or injected drugs. To construct the “injection network” (i.e. comprised of relationships in which individuals injected together), alters' names and demographic information were cross-referenced against that of study participants, with the assistance of community-based staff. UCINET [40] was used for network analysis and NetDraw [41] was used for network visualization.

2.2. Statistical analyses Evidence from research on vaccine uptake has demonstrated that intent is an important predictor of vaccine uptake [42–44]. However, meta-analyses have found that the average correlation between intention and behavior is often only moderate [45–47]. Given the debatable association between intent and behavior in similar vaccine research [48] and the desire to provide insight on

12

A.M. Young et al. / Contemporary Clinical Trials 41 (2015) 9–16

Table 2 Multivariate correlates to being ‘very likely’ to encourage drug-using partners to participate in a clinical trial for an HCV vaccine (n = 356).a Characteristicb Demographic Male Self-report of being in good health HCV-seropositive Total Monthly Income First quartile ($0–$200) Second quartile ($201–$698) Third quartile ($699–$1100) Fourth quartile (≥$1101) Substance use in past 6 months Heroin Methamphetamine Network characteristics Number of network members providing financial support

AOR (95% CI)

p-value

0.18 (0.08–0.37) 3.73 (1.70–8.16) 2.52 (1.29–4.92)

b0.001⁎ 0.001⁎ 0.007⁎ 0.034⁎

3.77 (1.51–9.40) 2.74 (1.08–6.97) 2.66 (1.04–6.81) Reference

3. Results Descriptive demographic and behavioral characteristics of the sample (n = 433) are described in detail elsewhere [57]. Most respondents were White (94%), male (55%), and unmarried (74%). The median age was 34 years (range: 21–68). Most (76%) reported injecting drugs in their lifetime, and 34% (n = 146) reported recent injection (past 6 months). Among recent injectors, 38% had shared injection equipment and a sizeable proportion had engaged in receptive and/or distributive needle sharing in the past 6 months (23% and 11%, respectively). 3.1. Attitudes toward clinical trial participation for an HCV vaccine

3.63 (0.80–16.50) 1.93 (0.57–6.49)

0.095 0.288

1.62 (1.05–2.49)

0.031⁎

AOR: adjusted odds ratio; CI: confidence interval. ⁎ p-value b 0.05. a Does not include those who reported that the question was not applicable (n = 18) or those who reported no network members (n = 59). b All variables were dichotomous with the exception of the continuous variable, “Number of network members providing financial support”.

behavior, a conservative dichotomization scheme was chosen (0 = Very unlikely/Unlikely/Likely; 1 = Very likely) in order to provide a more conservative estimate of future clinical trial participation. This coding scheme also addressed challenges posed by skewness in the response distributions. Hereafter, being ‘very likely’ to participate in a clinical trial is referred to as “willing to participate” (WTP) and being ‘very likely’ to encourage trial participation to drug-using partners is referred to as “willing to encourage participation” (WEP). Given that the target population for HCV vaccine trials will be HCV-seronegative PWID, WTP was assessed among HCVseronegative participants only and results specific to PWID are presented in the text. Aggregated data involving those who have and have not recently engaged in injection drug use were also analyzed, given that most had a history of injection drug use and relapse to injection is common among those who have ever injected [49,50]. Evidence suggests that these participants are nested in a dense risk network [39]; therefore, accounting for autocorrelation among participants' responses was required in analysis. Generalized linear mixed models, estimated using the PROC GLIMMIX [51] procedure (SAS software, v9.3) with a random effect for subject and Laplace approximation [52], were used for analyses. To account for possible biases presented by the use of respondent-driven sampling [53] to recruit the original cohort, all analyses were weighted using individualized weights based on individual network size computed using RDSAT 7.1 (Ithaca, NY) [54]. Odds ratios (ORs) and 95% confidence intervals (CIs) were reported. All variables that were significant (p-value b 0.05) in univariate analyses were evaluated in multivariate analysis. Collinearity in multivariate models was assessed using the %COLLIN_2011 macro [55]. Condition indexes of greater than 30 and corresponding variance decomposition proportions of greater than 0.5 were considered indicative of collinearity [56]. Collinearity was not present in either multivariate model.

Of the 433 participants, 266 (61%) were HCV-seropositive, and when queried about their willingness to participate in an HCV vaccine clinical trial, 158 reported that the question was not applicable because they were already infected with HCV (including two that were seronegative). Among the remaining participants (n = 165), 44% of the total sample (72/165) and 44% (33/75) of participants who had ever injected drugs reported that they would be ‘very likely’ to participate in a trial, and an additional 19% and 21%, respectively, indicated that they would be ‘likely’. Overall, 37% indicated that they would be ‘unlikely’ (18%) or ‘very unlikely’ (19%) to participate in an HCV vaccine clinical trial. Of the respondents who answered the question regarding WEP (n = 415, including 314 who had ever injected and 252 HCV-seropositive participants), the overwhelming majority were either ‘very likely’ (48%) or ‘likely’ (39%) to encourage their drug-using peers to participate in an HCV vaccine clinical trial. In the overall sample, few indicated that they were ‘unlikely’ (7%) or ‘very unlikely’ (6%) to do so. Of the 161 participants who answered the WEP and WTP questions, 23% (n = 15) of the 66 people who were WEP were not WTP, and of the people who were WTP (n = 71), just 72% were WEP. Many of the highly connected individuals in the injection network (i.e. in which ties represent injecting together) were not WTP or WEP in an HCV vaccine trial. Of the 47 injection network ties in which participants were willing to encourage trial participation, 17 (36%) would involve encouragement of an HCV-seropositive person. The serostatuses of the remaining 30 potential recipients of encouragement were unknown (i.e. they were not participants in the study). Of the 340 ‘isolates’ (i.e., participants reporting no injection partners), slightly fewer than half (47%) were WEP, and of the HCVseronegative isolates (n = 155), 46% were WTP. 3.2. Correlates to WTP and WEP in an HCV vaccine clinical trial Univariate analyses (Table 1) revealed that participants who were HCV-seropositive were significantly more likely to indicate WEP than those who were seronegative. Most demographic characteristics (e.g., age, race, education) were not associated with WEP; however, men were significantly less likely to indicate WEP, as were those with higher income. With the exception of heroin and methamphetamine use, behavioral characteristics were not associated with WEP. Participants who had used heroin and/or methamphetamine were more likely to report WEP, as were those who reported receipt of financial support from more network partners. When variables

A.M. Young et al. / Contemporary Clinical Trials 41 (2015) 9–16

significant in univariate analysis were entered into a multivariate model (Table 2), all with the exception of heroin and methamphetamine use remained significantly associated with WEP. Among HCV-seronegative participants (n = 165), lesser education, unemployment, receipt of financial support from more network members, and nonmedical use of prescription drugs were positively associated with WTP (Table 3). Other demographic and behavioral characteristics were not associated with WTP, although the negative association between male gender and WTP neared statistical significance (p-value = 0.064). In a multivariate model (Table 4) including only variables that were significant in univariate analysis, unemployment remained significantly associated with WEP. 4. Discussion In this sample of drug users, 63% of HCV-seronegative participants indicated that they would be very likely or likely to participate in a clinical trial for a prophylactic HCV vaccine, and 87% of the overall sample was either very likely or likely to encourage their peers to participate in HCV vaccine clinical trial research. This proportion reporting WTP is somewhat lower than that reported by other researchers, although comparisons are difficult given differences in survey instrumentation. In a study of young HCV-negative PWID from San Francisco, 88% of participants reported on a 4-point Likert scale item that they would be definitely (44%) or probably (44%) willing to participate in an HCV vaccine trial [21]. In a study of 113 Australian PWID, 74% indicated in response to an open-ended question that were willing to participate in an HCV vaccine trial [22]. Participants who were unemployed were significantly more likely to report willingness to participate in an HCV vaccine clinical trial. These findings are important given that differential enrollment by socioeconomic status could bias trial outcomes if behavioral or other clinically-relevant characteristics differ by socioeconomic status. Interestingly, characteristics traditionally related to unemployment, such as income and education, were not associated with WTP. Of note, the income measure used in this study includes income from all sources, including that made through illegal activities such as selling drugs; therefore, income in this study may not be as closely related to employment status as it would be in other populations. Education was associated with WTP in univariate analysis, but not after controlling for employment, drug use, and receipt of financial support from network members. Speculatively, unemployment may have emerged as having a stronger association with WTP than related socioeconomic indicators due to its relationship with other relevant factors, such as time availability and lack of access to private insurance; however, further research is needed to explore these possibilities. Previous research has indicated that financial incentives are likely to be key motivators of HCV vaccine trial participation among PWID [22–24]. Thus, as similar research has suggested [58], socioeconomic status of prospective participants must be taken into account when considering the use of monetary incentives during a clinical trial, as large incentives may induce coercion. Ensuring participants' comprehension of trial concepts is critical to the ethical conduct of HCV vaccine clinical trials, and research has indicated that provision of plain

13

language information coupled with brief oral discussion can significantly enhance comprehension of trial concepts [59]. An additional criterion for the ethical conduct of clinical trials is that participants have adequate access to resources promoting HCV prevention, such as substance abuse counseling, opiate substitution treatment (OST) and NSPs. Although Eastern Kentucky currently has clinics offering OST, particularly buprenorphine substitution [60], many opioid users in the region are uninsured or under-insured and the majority are unlikely to have the resources to pay for OST independently. In addition, Kentucky is one of 17 U.S. states with no NSPs [61], as such programs are currently proscribed by state law [35]. Thus, for the purposes of both research ethics and analysis of comparative efficacy, establishment of prevention opportunities must be integrated into the design and implementation of prophylactic HCV vaccine trials in Central Appalachia and other similarly underserved regions. Beyond the considerations of socioeconomic status, men were significantly less likely to indicate they would encourage trial participation among peers, but consistent with previous research [21], there were no gender differences with regard to WTP in a vaccine trial. Interestingly, similar research in this setting revealed that men were somewhat less willing to encourage HIV vaccination among peers but more likely to receive encouragement to get vaccinated [37]. Together, these findings highlight the instrumental role that women may play in disseminating positive vaccine messages among their drugusing peers. However, more research is needed to explore reasons for reluctance among men to encourage trial participation among peers and to examine individuals' receptivity to peer-promotion of trial participation. Participants who were HCV-seropositive were more than two times as likely to encourage HCV vaccine trial participation compared to participants who were HCV-seronegative. This finding indicates that HCV-seropositive individuals could play an important role in facilitating recruitment, particularly if they encourage trial participation in serodiscordant relationships involving risk behavior. However, many HCV-seropositive participants who indicated WEP were connected to other HCV-seropositive individuals in the injection network. In fact, the network data revealed that at least one in three potential ‘encouragements’ of trial participation in the injection network would be directed to an HCV-seropositive person. Thus, while network-based promotion may be a feasible strategy, efficiency should be evaluated as it may result in increased screening of HCV-seropositive individuals depending on the distribution of HCV in the network and the pattern of communication. Coupling a peer-driven approach with trial promotion by community organizations engaging HCV-seronegative PWID or practitioners aware of clients' HCV serostatuses may be the most efficient approach. The findings from this study should be generalized with caution and considered in light of study limitations, including reliance on self-reported behavioral data, focus on intent, and absence of measures to assess the influence of factors such as trial duration, side effects, and other trial-related characteristics. Though not ideal for assessment of theoretical constructs such as intent, one-item measures were used in an effort to minimize respondent-burden which is already elevated due to the network inventory and subsequent network memberspecific questions. Furthermore, predictive validity of the intent

14

A.M. Young et al. / Contemporary Clinical Trials 41 (2015) 9–16

Table 3 Univariate correlates to being ‘very likely’ to participate in a clinical trial for an HCV vaccine (n = 165).a Characteristicb

Willingness to participate in HCV vaccine trial Not very likelyc (n = 93)

Very likely (n = 72)

OR (95% CI)

p-value

55 (59.1) 87 (93.6) 36.6 (9.4)

30 (41.7) 69 (95.8) 37.0 (9.8)

0.42 (0.17–1.05) 2.06 (0.28–15.23) 1.01 (0.97–1.06)

0.064 0.479 0.516 0.133

22 (23.6) 23 (24.7) 19 (20.4) 29 (31.2) 11.7 (1.7) 28 (30.1) 60 (64.5) 21 (22.6) 59 (63.4) 6 (6.5)

18 (25.0) 18 (25.0) 23 (31.9) 13 (18.1) 11.0 (2.1) 29 (40.3) 42 (58.3) 29 (40.3) 51 (70.8) 5 (6.9)

3.49 (1.02–11.90) 2.88 (0.88–9.43) 3.96 (1.12–13.99) Reference 0.78 (0.61–0.99) 1.89 (0.77–4.62) 0.67 (0.28–1.57) 3.78 (1.51–9.45) 1.07 (0.46–2.48) 0.61 (0.11–3.41)

0.044⁎ 0.162 0.354 0.004⁎⁎ 0.884 0.570

Substance use in past 6 months Alcohol — n(%) Alcohol to intoxication — n(%) Nonmedical use of prescription drugs — n(%) Heroin — n(%) Methamphetamine — n(%) Cocaine — n(%) Crack — n(%)

41 (44.1) 32 (34.4) 74 (79.6) 1 (1.1) 6 (6.5) 8 (8.6) 4 (4.3)

33 (45.8) 20 (27.8) 66 (91.7) 4 (5.6) 7 (9.7) 8 (11.1) 1 (1.4)

1.20 (0.53–2.70) 0.93 (0.37–2.33) 4.11 (1.24–13.69) 3.44 (0.30–39.85) 2.23 (0.44–11.21) 1.39 (0.30–6.46) 0.14 (0.00–5.90)

0.664 0.881 0.021⁎ 0.322 0.331 0.674 0.306

Risk behaviord Injected drugs — n(%) Injected with unclean needle — n(%) Distributed unclean needlee — n(%) Shared injection equipmentf — n(%) Bleached needles — n(%) Unprotected sex with PWID — n(%) Snorted drugs (past 30 days) — n(%)

8 (8.6) 2 (2.2) 1 (1.1) 3 (3.2) 3 (3.2) 8 (8.6) 53 (57.0)

9 (12.5) 2 (2.8) 0 (0.0) 2 (2.8) 1 (1.4) 5 (6.9) 38 (52.8)

1.65 (0.43–6.32) 0.63 (0.02–18.96) – 0.36 (0.02–7.87) 0.35 (0.02–7.37) 0.49 (0.10–2.33) 0.67 (0.30–1.51)

0.464 0.793 – 0.513 0.501 0.371 0.337

0.11 (0.52) 0.29 (0.80) 0.27 (0.94)

0.15 (0.81) 0.29 (0.91) 0.16 (0.55)

1.18 (0.69–2.01) 1.04 (0.67–1.60) 0.73 (0.38–1.37)

0.539 0.878 0.325

0.14 (0.83)

0.35 (2.16)

1.01 (0.70–1.47)

0.954

1.46 (1.79) 3.94 (5.19) 1.75 (1.32) 0.59 (0.73)

1.10 (1.31) 2.71 (3.28) 1.85 (1.32) 0.95 (1.02)

0.87 (0.66–1.15) 0.94 (0.84–1.04) 1.05 (0.76–1.46) 1.75 (1.01–3.03)

0.334 0.234 0.768 0.047⁎

Demographic Male — n(%) White — n(%) Age — mean (SD) Total monthly income First quartile ($0–$200) Second quartile ($201–$698) Third quartile ($699–$1100) Fourth quartile (≥$1101) Education (years) — mean (SD) Married — n(%) Uninsured — n(%) Unemployed — n(%) Self-report of being in good health — n(%) Incarcerated (past 30 days) — n(%)

Network characteristics (n = 145)g Number of injection partners — mean (SD) Number of people in personal network who inject — mean (SD) Number of people to whom participant talks about injection-related risk reduction — mean (SD) Frequency of sharing injection equipment with all network members — mean (SD) Number of people with whom participant uses drugs – mean (SD) Total frequency of drug sharing with all partners — mean (SD) Number of network members providing social support – mean (SD) Number of network members providing financial support — mean (SD)

PWID: person who injects drugs; OR: odds ratio; CI: confidence interval; SD: standard deviation. ⁎ p-value b 0.05. ⁎⁎ p-value b 0.01. a Does not include 266 participants who were already HCV positive (confirmed using OraQuick Rapid-HCV tests) and two additional participants who were prompted to skip WTP questions when they self-reported being HCV positive although confirmatory testing had not yet been performed. b All categorical variables (indicted with a ‘n(%)’) were dichotomous. c Includes responses “very unlikely”, “unlikely”, and “likely”. d The recall period for risk behaviors listed was 6 months with the exception of snorting, which was assessed based on the past 30 days. e Sold, loaned, or gave needle to someone after using it. f Cookers, cottons, and/or rinse water. g 20 participants reported no network members.

measure can only be assessed in the context of an available clinical trial (i.e., assessing intent, then offering opportunity to enroll and examining correspondence). Of note, this study was conducted in a rural population that, despite its elevated HCV burden, has been significantly underrepresented in similar research to date. The findings based on this sample, comprised predominately of white, nonmedical prescription drug users, may not be generalizable to urban drug-using populations with different demographic, behavioral, and network characteristics.

In this study, most participants who had recently (past 6 months) engaged in injection drug use were HCV-seropositive, leaving only 17 HCV-seronegative PWID for inclusion in the WTP analysis and precluding our ability to evaluate correlates of WTP specifically among PWID. However, the aggregated analyses including those who had not injected in the past 6 months remain valuable given that most had a history of injection and could be prone to re-initiation of injection and HCV acquisition. Also, this study did not assess

A.M. Young et al. / Contemporary Clinical Trials 41 (2015) 9–16 Table 4 Multivariate correlates to being ‘very likely’ to participate in a clinical trial for an HCV vaccine (n = 145).a Characteristicb

AOR (95% CI)

p-value

Demographic Education (years) Unemployed

0.85 (0.66–1.09) 4.52 (1.59–12.86)

0.189 0.005⁎

Substance use in past 6 months Nonmedical use of prescription drugs

3.97 (0.96–16.49)

0.058

1.68 (0.95–2.96)

0.074

Network characteristics Number of network members providing financial support

AOR: adjusted odds ratio; CI: confidence interval. ⁎ p-value b 0.05. a Does not include those who reported no network members (n = 20), those who were already HCV positive (confirmed using OraQuick Rapid-HCV tests, n = 266), and two additional participants who were prompted to skip WTP questions when they self-reported being HCV positive although confirmatory testing had not yet been performed. b All variables were dichotomous with the exception of the continuous variable, “Number of network members providing financial support”.

straw sharing, a behavior that should be examined in future research given its association with HCV transmission [18,19]. In addition, participants were asked only about their WEP among drug-using peers in general and not on a partner-by-partner basis. Given previous research that suggests selectivity in vaccine communication among social, sexual and drug network members [37], future studies should examine not only if participants will encourage trial participation but to whom. Nevertheless, it is notable that indication of strong willingness to encourage trial participation among peers was present in nearly half the sample. Despite evidence that peer communication and support can play a role in promoting participation in HCV clinical trial research [20,23], no study to date has explored feasibility of involving peer-promotion in HCV vaccine trial recruitment among drug users. This study's findings suggest that a peerdriven approach to recruitment for a prophylactic HCV vaccine trial could be possible to implement in this population and has potential to reach a large number of individuals in the community at risk for acquiring HCV. However, several notable findings emerged. This study revealed that selecting “peer promoters” based on their centrality in the local injection network, in terms of their number of contacts, may not be the most efficient approach; many participants with multiple partners in the risk network reported being unlikely to encourage trial participation among their peers. Moreover, many of the partners of those who were central in the injection network were already HCV positive and would not benefit from a prophylactic vaccine. In this setting, women and HCVseropositive individuals would be more appropriate promoters, as they indicated the most willingness to encourage others to enroll in a trial. Moreover, HCV-seropositive individuals are likely connected to those who are at high risk for incident infection. This study also revealed that individuals of lower socioeconomic status may be more willing to participate in clinical trials, indicating that drug users of higher socioeconomic status may be the hardest to reach in this ‘hard-to-reach’ population. This finding underscores the need for more formative research into trial logistics, incentive structure, and messaging that could promote enrollment across the

15

socioeconomic spectrum. Overall, these findings indicate that while a network-based approach may be an effective strategy for trial recruitment and vaccine promotion, the strategy should be coupled with other approaches such as social marketing or public endorsement by community leaders. Finally, safeguards to prevent the coercion of prospective lowincome individuals along with the provision of established HCV preventive services for enrolled study participants are crucial factors to consider in the ethical design of vaccine trial research in this and similar at-risk populations. Acknowledgments The Social Networks among Appalachian People Study is funded by the National Institute on Drug Abuse (R01DA024598 and R01DA033862 to J.R.H). Data collection on attitudes surrounding HCV vaccine clinical trial participation was supported by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health (UL1TR000117 to J.R.H) and D.B.S. was supported by TL1RR033172. References [1] Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair-Rohani H, et al. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380(9859):2224–60 [Epub 2012/12/19]. [2] Ly KN, Xing J, Klevens RM, Jiles RB, Ward JW, Holmberg SD. The increasing burden of mortality from viral hepatitis in the United States between 1999 and 2007. Ann Intern Med 2012;156(4):271–8 [Epub 2012/02/22]. [3] Ward JW. The epidemiology of chronic hepatitis C and one-time hepatitis C virus testing of persons born during 1945 to 1965 in the United States. Clin Liver Dis 2013;17(1):1–11 [Epub 2012/11/28]. [4] Centers for Disease Control and Prevention. Use of enhanced surveillance for hepatitis C virus infection to detect a cluster among young injectiondrug users—New York, November 2004–April 2007. MMWR Morb Mortal Wkly Rep 2008;57(19):517–21 [Epub 2008/05/16]. [5] Centers for Disease Control and Prevention. Hepatitis C virus infection among adolescents and young adults: Massachusetts, 2002–2009. MMWR Morb Mortal Wkly Rep 2011;60(17):537–41 Epub 2011/05/06. [6] Williams IT, Bell BP, Kuhnert W, Alter MJ. Incidence and transmission patterns of acute hepatitis C in the United States, 1982–2006. Arch Intern Med 2011;171(3):242–8 [Epub 2011/02/18]. [7] Centers for Disease Control and Prevention. Updated U.S. public health service guidelines for the management of occupational exposures to HBV, HCV, and HIV and recommendations for postexposure prophylaxis. MMWR Recomm Rep 2001;50(RR-11):1–52 [Epub 2001/07/10]. [8] Page-Shafer K, Hahn JA, Lum PJ. Preventing hepatitis C virus infection in injection drug users: risk reduction is not enough. AIDS 2007;21(14): 1967–9 [Epub 2007/08/28]. [9] Hahn JA, Wylie D, Dill J, Sanchez MS, Lloyd-Smith JO, Page-Shafer K, et al. Potential impact of vaccination on the hepatitis C virus epidemic in injection drug users. Epidemics 2009;1(1):47–57 [Epub 2009/03/01]. [10] Krahn MD, John-Baptiste A, Yi Q, Doria A, Remis RS, Ritvo P, et al. Potential cost-effectiveness of a preventive hepatitis C vaccine in high risk and average risk populations in Canada. Vaccine 2005;23(13):1549–58 [Epub 2005/02/08]. [11] Massad E, Coutinho FA, Chaib E, Burattini MN. Cost-effectiveness analysis of a hypothetical hepatitis C vaccine compared to antiviral therapy. Epidemiol Infect 2009;137(2):241–9 [Epub 2008/07/18]. [12] Torresi J, Johnson D, Wedemeyer H. Progress in the development of preventive and therapeutic vaccines for hepatitis C virus. J Hepatol 2011; 54(6):1273–85 [Epub 2011/01/18]. [13] Fauvelle C, Lepiller Q, Felmlee DJ, Fofana I, Habersetzer F, Stoll-Keller F, et al. Hepatitis C virus vaccines—progress and perspectives. Microb Pathog 2013;58:66–72 [Epub 2013/03/19]. [14] Swadling L, Klenerman P, Barnes E. Ever closer to a prophylactic vaccine for HCV. Expert Opin Biol Ther 2013;13(8):1109–24 [Epub 2013/05/09]. [15] Law LMJ, Landi A, Magee WC, D. Lorne T, Houghton M. Progress towards a hepatitis C virus vaccine. Emerg Microbes Infect 2013;2:e79.

16

A.M. Young et al. / Contemporary Clinical Trials 41 (2015) 9–16

[16] Cox AL, Thomas DL. Hepatitis C virus vaccines among people who inject drugs. Clin Infect Dis 2013;57(Suppl. 2):S46–50 [Epub 2013/08/02]. [17] White B, Madden A, Prins M, Hellard M, Wand H, Dore GJ, et al. Assessing the feasibility of hepatitis C virus vaccine trials: results from the Hepatitis C Incidence and Transmission Study-community (HITS-c) vaccine preparedness study. Vaccine 2014;32(42):5460–7 [Epub 2014/08/19]. [18] Aaron S, McMahon JM, Milano D, Torres L, Clatts M, Tortu S, et al. Intranasal transmission of hepatitis C virus: virological and clinical evidence. Clin Infect Dis 2008;47(7):931–4. [19] Macias J, Palacios RB, Claro E, Vargas J, Vergara S, Mira JA, et al. High prevalence of hepatitis C virus infection among noninjecting drug users: association with sharing the inhalation implements of crack. Liver Int 2008;28(6):781–6. [20] Maher L, White B, Hellard M, Madden A, Prins M, Kerr T, et al. Candidate hepatitis C vaccine trials and people who inject drugs: challenges and opportunities. Vaccine 2010;28(45):7273–8 [Epub 2010/09/14]. [21] Levy V, Evans JL, Stein ES, Davidson PJ, Lum PJ, Hahn JA, et al. Are young injection drug users ready and willing to participate in preventive HCV vaccine trials? Vaccine 2010;28(37):5947–51 [Epub 2010/07/20]. [22] Park JN, White B, Bates A, Enriquez J, Liao L, Maher L. Motivators and barriers influencing willingness to participate in candidate HCV vaccine trials: perspectives of people who inject drugs. Drug Alcohol Depend 2012;123(1–3):35–40 [Epub 2011/11/11]. [23] Treloar C, Byron P, McCann P, Maher L. “Fitness for duty”: social, organisational and structural influences on the design and conduct of candidate hepatitis C vaccine trials involving people who inject drugs. Vaccine 2010;28(32):5228–36 [Epub 2010/06/12]. [24] Maher L, White B, Donald A, Bates A, Enriquez J, Pham S, et al. Using ethnographic fieldwork to inform hepatitis C vaccine preparedness studies with people who inject drugs. Int J Drug Policy 2010;21(3): 194–201 [Epub 2009/06/02]. [25] Doab A, Topp L, Day CA, Dore GJ, Maher L. Clinical trial literacy among injecting drug users in Sydney, Australia: A pilot study. Contemp Clin Trials 2009;30(5):431–5 [Epub 2009/04/21]. [26] Valente TW, Zogg JB, Christensen S, Richardson J, Kovacs A, Operskalski E. Using social networks to recruit an HIV vaccine preparedness cohort. J Acquir Immune Defic Syndr 2009;52(4):514. [27] Havens JR, Lofwall MR, Frost SD, Oser CB, Leukefeld CG, Crosby RA. Individual and network factors associated with prevalent hepatitis C infection among rural Appalachian injection drug users. Am J Public Health 2013;103(1):e44–52 [Epub 2012/11/17]. [28] Christian WJ, Hopenhayn C, Christian A, McIntosh D, Koch A. Viral hepatitis and injection drug use in Appalachian Kentucky: a survey of rural health department clients. Public Health Rep 2010;125(1):121. [29] Centers for Disease Control and Prevention. Viral hepatitis statistics & surveillance. [cited 2014 February 21]; Available from: http://www.cdc. gov/hepatitis/Statistics/2011Surveillance/Table4.1.htm; 2013. [30] Havens J. Injection drug use and other risk behaviors among rural prescription opioid users. The College on Problems of Drug Dependence Annual Meeting; June 2014; San Juan, Puerto Rico; 2014. [31] White B, Madden A, Prins M, Hellard M, Wand H, Dore GJ, et al. Assessing the feasibility of hepatitis C virus vaccine trials: results from the Hepatitis C Incidence and Transmission Study-community (HITS-c) vaccine preparedness study. Vaccine 2014;32(42):5460–7. [32] Appalachian Regional Commission. An analysis of mental health and substance abuse disparities & access to treatment services in the Appalachian region. [updated August 2008; cited 2013 August 15]; Available from: http://www.arc.gov/research/researchreportdetails.asp? REPORT_ID=71; 2008. [33] Halverson J. An analysis of disparities in health status and access to health care in the Appalachian Region; 2004. [34] Appalachian Regional Commission, Economic overview of Appalachia. Regional Planning and Research Division. 2011 [updated January 28, 2011; cited 2014 July 8]; Available from: http://www.arc.gov/images/appregion/ Jan2011/EconomicOverview-1-28-11.pdf; 2010. [35] Kentucky Legislative Research Commission. Definitions for KRS 218A.500 and 218A.510 — Unlawful practices — Penalties 2010. [updated April 26, 2010; cited 2013 May 10]; Available from: http://www.lrc.ky.gov/krs/ 218a00/500.PDF; 2010. [36] Kentucky Legislative Research Commission. Sale and disposal of hypodermic syringes or needles, KRS 217.177. [updated June 20, 2005; cited 2013 May 10]; Available from: http://www.lrc.ky.gov/krs/217-00/177.pdf; 2005.

[37] Young AM, DiClemente RJ, Halgin DS, Sterk CE, Havens JR. Drug users' willingness to encourage social, sexual, and drug network members to receive an HIV vaccine: a social network analysis. AIDS Behav 2014;18: 1753–63. [38] Cha YJ, Park Q, Kang ES, Yoo BC, Park KU, Kim JW, et al. Performance evaluation of the OraQuick hepatitis C virus rapid antibody test. Ann Lab Med 2013;33(3):184–9 [Epub 2013/05/15]. [39] Young A, Jonas A, Mullins U, Halgin D, Havens J. Network structure and the risk for HIV transmission among rural drug users. AIDS Behav 2013;17(7): 2341–51. [40] Borgatti S, Everett M, Freeman L. Ucinet for Windows: software for social network analysis. 6.303 ed. Harvard, MA: Analytic Technologies; 2002. [41] Borgatti S. Netdraw network visualization. 2.095 ed. Harvard, MA: Analytic Technologies; 2002. [42] Painter JE, Sales JM, Pazol K, Wingood GM, Windle M, Orenstein WA, et al. Adolescent attitudes toward influenza vaccination and vaccine uptake in a school-based influenza vaccination intervention: a mediation analysis. J Sch Health 2011;81(6):304–12. [43] Liao Q, Cowling BJ, Wendy Wing Tak L, Richard F. Factors affecting intention to receive and self-reported receipt of 2009 pandemic (H1N1) vaccine in Hong Kong: a longitudinal study. PLoS One 2011;6(3):1–13. [44] Patel DA, Zochowski M, Peterman S, Dempsey AF, Ernst S, Dalton VK. Human papillomavirus vaccine intent and uptake among female college students. J Am Coll Health 2012;60(2):151–61. [45] Sheppard BH, Hartwick J, Warshaw PR. The theory of reasoned action: a meta-analysis of past research with recommendations for modifications and future research. J Consum Res 1988;15:325–43. [46] Armitage C, Conner M. Efficacy of the theory of planned behavior: a metaanalytic review. Br J Soc Psychol 2001;40(29):471–99. [47] Randall D, Wolff J. The time interval in the intention behaviour relationship: meta analysis. Br J Soc Psychol 1994;33(4):405–18. [48] Poole G. Using psychological principles to narrow the intention–behavior gap and increase participation in HIV vaccine trials. Curr HIV Res 2012; 10(6):552–6. [49] Evans JL, Hahn JA, Lum PJ, Stein ES, Page K. Predictors of injection drug use cessation and relapse in a prospective cohort of young injection drug users in San Francisco, CA (UFO Study). Drug Alcohol Depend 2009;101(3):152–7. [50] Shah NG, Galai N, Celentano DD, Vlahov D, Strathdee SA. Longitudinal predictors of injection cessation and subsequent relapse among a cohort of injection drug users in Baltimore, MD, 1988–2000. Drug Alcohol Depend 2006;83(2):147–56. [51] SAS Institute. The GLIMMIX Procedure. SAS/STAT 93 user's guide. Cary, NC: SAS Institute; 2011. [52] Wolfinger R. Laplace's approximation for nonlinear mixed models. Biometrika 1993;80(4):791–5. [53] Heckathorn D. Respondent-driven sampling II: deriving valid population estimates from chain-referral samples of hidden populations. Soc Probl 2002;49(1):11–34. [54] Volz E, Wejnert C, Cameron C, Spiller M, Barash V, Degani I, et al. Respondent-driven sampling analysis tool (RDSAT) version 7.1. Ithaca, NY: Cornell University; 2012. [55] Zack M, Singleton J, Satterwhite C, Delaney K, Wall K. Collinearity macro (SAS) (contact David Kleinbaum at [email protected]); 2011. [56] Kleinbaum D, Klein M. Logistic regression: a self-learning text. 3rd ed. New York: Springer; 2010. [57] Young AM, DiClemente RJ, Halgin DS, Sterk CE. HIV vaccine acceptability among high-risk drug users in Appalachia: a cross-sectional study. BMC Public Health 2014;14(537). [58] Fisher CB. Enhancing HIV, vaccine trial consent preparedness among street drug users. J Empir Res Hum Res Ethics 2010;5(2):65–80 [Epub 2010/06/ 24]. [59] White B, Madden A, Hellard M, Kerr T, Prins M, Page K, et al. Increased hepatitis C virus vaccine clinical trial literacy following a brief intervention among people who inject drugs. Drug Alcohol Rev 2013;32(4):419–25 [Epub 2012/11/02]. [60] Substance Abuse and Mental Health Services Administration. Behavioral health treatment services locator. [updated 2014; cited 2014 October 6]; Available from: http://findtreatment.samhsa.gov/; 2014. [61] Foundation for AIDS Research. Syringe Services Program Coverage in the United States. [cited 2014 July 1]; Available from: http://www.amfar.org/ uploadedFiles/_amfarorg/Articles/On_The_Hill/2013/2013%20SSP% 20Map%20Final.pdf; 2013.